Table 2.
Predicted protein level consequences and potential pathogenicity of rare variants in study family.
| Predicted change at protein level due to rare variant | Poly Phen-2 | phyloP | MutationTaster | PROVEAN | CENTO-GENE | Known clinical phenotypes associated with other variation in gene | Likelihood of pathogenicity based on ACMG guidelines |
|---|---|---|---|---|---|---|---|
| MLKL: Frameshift deletion: 369-471del + novel 21 a/a | N/A | N/A | N/A | N/A | N/A | Heterozygous variant leading to reduced expression associated with late-onset Alzheimer’s disease (see main manuscript for reference) | Moderate evidence of pathogenicity |
| FA2H: Non-frame-shift deletion: F11del | N/A | N/A | N/A | −1.305 (neutral) | Cannot be confirmed as damaging | Multiple FA2H variants associated with hereditary spastic paraplegia, leukodystrophy, neurodegeneration with brain iron accumulation (see main manuscript for references) | Moderate evidence of pathogenicity |
| AP1S2: Non-synonymous single nucleotide variant: K55R | 0.642 (poss.) | 0.99847 (neutral) | 0.96041 (neutral) | −2.92 (poss.) | N/A | Variants causing substantial protein truncation or deletion associated with mental retardation, hypotonia, microcephaly, developmental delays with age of onset at 4 years or less | Supporting evidence of benign impact (clinical phenotype mismatch) |
The predicted protein level consequences and potential pathogenicity of the rare variants shown in Supplementary Table 1, prior to functional investigations. Potential pathogenicity was assessed using in silico predictive algorithms where possible; these algorithms generate scores for the likelihood of the variant affecting protein function (where ‘poss.’ denotes a possibly damaging effect and ‘neutral’ indicates no likely effect). The FA2H non-frameshift deletion F11del was also evaluated by CENTOGENE (www.centogene.com), a CAP accredited company for clinical genetic testing, who could not confirm the variant as damaging. Known clinical phenotypes for other variants in the genes in question are also summarized. The likelihood of variant pathogenicity as estimated by the ACMG guidelines considers variant frequency, potential protein level impact, familial disease pedigree, other known variants in the gene of interest, and the clinical phenotypes they contribute to.